Osteosarcoma is a highly aggressive primary bone malignancy whose diagnosis and treatment planning heavily rely on histopathology slide-level analysis. Manual analysis of histopathology slides is very time-consuming and also requires skilled pathologists to examine them. Recent advances in deep learning have demonstrated outstanding performance in digital pathology. However, most Convolutional Neural Network (CNN) approaches rely heavily on global pooling, which is unable to capture diagnostically critical areas. This study proposes an attention pooling-based patch-level classification for osteosarcoma histopathology images. The proposed model consists of two modules: The first is a hierarchical feature extractor that uses ConvNeXt tiny as its backbone for feature extraction, and the second is an attention pooling model that assigns adaptive weights to emphasize spatial importance within the deep feature maps. It allows the model to concentrate on diagnostically relevant areas. The experimental results demonstrate that the attention pooling-based feature flattening utilized in the proposed framework improves the performance. This proposed model is evaluated on an osteosarcoma tumor assessment dataset. The proposed model outperforms the existing models with an accuracy of 92.31%.

Fetal phonocardiography (FPCG) is a non-invasive Fetal Heart Rate (FHR) monitoring technique that can detect vibrations and murmurs in heart sounds. However, acquiring fetal heart sounds from a wearable FPCG device is challenging due to noise and artefacts. This research contributes a resilient solution to overcome the conventional issues by adopting Artificial Intelligence (AI) with FPCG for automated FHR monitoring in an end-to-end manner, named (AI-FHR). Four sequential methodologies were used to ensure reliable and accurate FHR monitoring. The proposed method removes low-frequency noises and high-frequency noises by using Chebyshev II high-pass filters and Enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ECEEMDAN) in combination with Phase Shifted Maximal Overlap Discrete Wavelet Transform (PS-MODWT) filters, respectively. The denoised signals are segmented to reduce complexity, and the segmentation is performed using multi-agent deep Q-learning (MA-DQL). The segmented signal is provided to reduce the redundancies in cardiac cycles using the Artificial Hummingbird Optimization (AHBO) algorithm. The segmented and non-redundant signals are converted into 3D spectrograms using a machine learning algorithm called variational auto-encoder-general adversarial networks (VAE-GAN). The feature extraction and classification are carried out by adopting a hybrid of the bidirectional gated recurrent unit (BiGRU) and the multi-boosted capsule network (MBCapsNet). The proposed method was implemented and simulated using MATLAB R2020a and validated by adopting effective validation metrics. The results demonstrate that the proposed method performed better than the current method with accuracy (81.34%), sensitivity (72%), F1-score (83%), Energy (0.808 J), and complexity index (13.34). Like other optimization methods, AHO needs precise parameter adjustment in order to function well. Its performance may be greatly impacted by the selection of parameters, including population size, exploration rate, and learning rate.

Parkinson's disease (PD), a harmful scenario that decreases the value of lifestyle. Those who have this disease are having difficulty in writing, speaking, and walking. According to some research speech analysis is the best technique used to detect the PD because majority of people have speech disorders who are suffering from PD. Detecting the disease in its early stages is the challenge so that it does not get worse. For this, we are using machine learning algorithms for the classification of PD detection. The various classification models like support vector machines, Logistic Regression, KNN, and random forest are effectively used for classification purposes. By using different classification models, we can classify them and predict the accuracy, compare them with other models, and see which best fits for classifying/detecting PD. We are using a dataset in which there are some records based on the voice signals of individuals which helps us to detect who has Parkinson's or not. Machine learning is very good at recognizing patterns and can identify patterns in data that can help with analysis. We are using different metric calculations such as finding out the precision, f1_score, recall, and confusion matrix as well which gives us an idea about the designed models that we are using, and also accurate results so that can be used for detecting PD.

Cardiovascular diseases are some of the most common diseases today. A new estimate from the World Heart Federation (WHF) states that the number of deaths from cardiovascular diseases (CVD) increased from 12.1 million in 1990 to 20.5 million in the year 2021. In recent years, the field of healthcare has witnessed significant advancements in technology and data analysis techniques. Congenital abnormalities, diseases caused by impaired heart rhythm, vascular occlusion, post-operation arrhythmias, heart attacks and irregularities in heart valves are some of the various cardiovascular diseases. Early recognition of them is very important for obtaining positive results in treatment. One such area of research that holds great promise is the classification of heart sounds using phonocardiography (PCG). Classification of heart sounds has become increasingly important in enhancing diagnostic precision and enhancing patient care. The integration of machine and deep learning into medical diagnostics has emerged as a transformative avenue. Machine and Deep learning techniques offer the potential to automate and increase the accuracy of cardiac sound analysis providing healthcare professionals with rapid and constant diagnostics support. This research contributes to efficient cardiac diagnostics, aiding the timely detection of abnormalities and enhancing patient care. It aspires to learn about the value of heart sound classification, the utilization of phonocardiography, and the application of deep learning and machine learning methods for enhancing the accuracy of classification.

This study explores advanced Electroencephalogram (EEG)-based methods for epileptic seizure detection, utilizing preprocessing and diverse feature extraction techniques. Employing machine learning algorithms like Random Forest, it achieves promising accuracy, offering valuable insights for timely intervention and personalized treatment in epilepsy management, contributing to enhanced diagnostic tools.

In general, hospital procedures require a large number of medical reports, which are an essential component of the process. Hospitals are now increasing their business by integrating pathology and other test laboratories into the hospital for efficient and speedy reporting, as well as increased income. Hospital operations cover a wide range of functions, from patient admission and management to hospital expenditure management. This, together with additional services such as pathology and pharmacy management, adds operational complexity and makes it harder to track. To address this issue, we employ blockchain technology to keep track of every single transaction with 100% authenticity using the hyperledger idea. All transactions are encrypted and saved as blocks to enable authentication over a network of computers rather than a centralized server. Further-more, we employ the hyperledger idea to associate and preserve all of the medical papers linked with each transaction, including the date stamp. This enables for the authentication of each report, which will be identified if edited by anybody.

Among diabetic patients, Diabetic Retinopathy (DR) is one of the main causes of blindness; therefore, early and accurate detection is essential for successful treatments. Convolutional Neural Network, one type of deep learning technique, has demonstrated potential in automating the diagnosis of diabetic retinal disease using retinal pictures. We provide a new method in this paper for detecting diabetic retinopathy that makes use of the Inception Net architecture. Because of its reputation for processing high-resolution images efficiently, the Inception Net model is a good fit for the intricate tasks involved in retinal image analysis. We trained and assessed our proposed model using a large dataset of annotated retinal pictures, and it achieved high specificity, sensitivity, and accuracy in differentiating between retinas that were healthy and those that were diseased. According to our research, deep learning-based methods like Inception Net have a great deal of promise for the accurate and fast identification of diabetic retinopathy, which will lead to better patient outcomes and enable prompt clinical intervention.

Analysis of Phonocardiogram (PCG) data is crucial in the diagnosis of cardiovascular conditions. This research introduces an innovative method for automatically categorizing heart sounds into five distinct groups: murmurs, artifacts, extrasystole, extrahls, and normal sounds. Advanced machine learning techniques are used to extract Mel-Frequency Cepstral Coefficients (MFCCs) from PCG signals as discerning features. Augmentation methods are employed to increase the training dataset, thereby enhancing the model's generalization capability. The classification is carried out using the K- Nearest Neighbors (KNN) algorithm, which achieves an impressive 91% accuracy across the specified categories. The developed framework showcases the effectiveness of machine learning in automating heart sound analysis, leading to improved diagnostic precision and efficiency. The reproducible nature of the provided code enables wider adoption and facilitates further research in this domain. This work contributes to advancing cardiac diagnostics, offering valuable insights for both clinical practice and research in cardiovascular health.

India is a developing country with a rapidly in-creasing population. In terms of population, India is ranked second in the globe. The main motive is to provide a smart traffic signal for emergency vehicles. Each vehicle has a unique radio frequency identification (RFID) tag that makes it difficult to remove or destroy. One of the most difficulties that metro areas confront these days is traffic monitoring and management. We have proposed a mechanism in this manuscript to dynamically arrange traffic lights to eliminate traffic congestion and allow emergency vehicles to move freely on the route. Existing concepts include using timers for each phase of a traffic light or using electronic sensors to identify vehicles. When he notices the ambulance, the other option is to enlist the assistance of traffic cops. Proposed method uses an Arduino UNO microcontroller, the reader module, and a RFID tag to design the methodology in this manuscript. The RFID reader detects a RFID tag in its vicinity and sends the RFID tag number to the Arduino. Inside the RFID Tag is a coil and a chip. When this ID is in close contact to the scanner, electromagnetic induction induces electricity in the coil, which lights up the chip. When the ambulance or any emergency vehicle uses RFID tag to pass through signals, the signal turns green, while all other signals at the intersection remain red. This allows the ambulance or any emergency vehicle to pass without having to wait for a green signal in densely populated or congested locations.

With the fast-growing economy in which the majority of the workforce uses two-wheelers, the occurrence of accidents has increased by 35 percent over a 35-year period, with fatalities totaling around 58,000 last year. The main cause of accidents is that the rider does not follow safety protocols or their accident is not reported on time. We proposed a smart helmet that detects accidents and detects if the rider is intoxicated by alcohol when worn by the rider. The prototype uses the following sensors to detect this (IR Sensor, Accelerometer, Breath-analyzer). The accelerometer measures the rider's sudden change in tilt and sends data to a programmed interface. The breathalyser will detect the amount of alcohol in the rider's breath and report if the reading exceeds the legal limit. The server gathers the information from the IR sensor to train Support Vector Machine (SVM) [1] which will be useful to optimize accident detection in the future when sufficient data is gathered.

SARS-CoV-2 started a global epidemic that resulted in COVID-19, a real infectious disease that disrupted regular living all over the world. Sterilizing our hands is crucial since the virus and other diseases are spread by touching contaminated surfaces. In this manuscript, a prototype for low-cost sterilisation is created that uses an IR thermal sensor to measure temperature and UV C light rays to disinfect our hands. Numerous bacteria are affected throughout the sanitization process, which has a number of advantages over chemical-based sanitization techniques. In contrast to relevant, it is also easy to customise. There are proprietary devices that can be purchased commercially. This gadget is an excellent illustration of open-source technology. automatic, quick, and safe hand sanitising device.

Bacteria such as Escherichia coli exhibit biased motion if kept in a spatially nonuniform chemical environment. Here, we bring out unique time-dependent characteristics of bacterial chemotaxis, in response to a diffusing spatial step ligand profile. Experimentally obtained temporal characteristics of the drift velocity are compared with the theoretical and Monte Carlo simulation-based estimates, and excellent agreements can be obtained. These results bring insights to the time-responsive facets of bacterial drift, bearing far-reaching implications in understanding their migratory dynamics in the quest of finding foods by swimming toward the highest concentration of food molecules, or for fleeing from poisons, as well as toward the better understanding of therapeutic response characteristics for certain infectious diseases.

Bacterial species are known to show chemotaxis, i.e., the directed motions in the presence of certain chemicals, whereas the motion is random in the absence of those chemicals. The bacteria modulate their run time to induce chemotactic drift towards the attractant chemicals and away from the repellent chemicals. However, the existing theoretical knowledge does not exhibit a proper match with experimental validation, and hence there is a need for developing alternate models and validating experimentally. In this paper a more robust theoretical model is proposed to investigate chemotactic drift of peritrichous Escherichia coli under an exponential nutrient gradient. An exponential gradient is used to understand the steady state behavior of drift because of the logarithmic functionality of the chemosensory receptors. Our theoretical estimations are validated through the experimentation and simulation results. Thus, the developed model successfully delineates the run time, run trajectory, and drift velocity as measured from the experiments.

In this paper, a new method, directional line edge binary pattern (DLEBP) is proposed for image indexing and retrieval. This method is the extension of line edge binary pattern (LEBP). The directional line edges of neighbours for a particular center pixel are determined by eight directional windows. The 3x3 directional window is considered for this purpose. The retrieval results of the proposed method are tested on Brodatz image database. The performance of the proposed method is measured in terms of average retrieval rate (ARR). The results after being investigated show a significant improvement in terms of ARR as compared to Gabor transform (GT), dual-tree complex wavelet transform (DTCWT), dual-tree rotated complex wavelet transform (DT-RCWT), local ternary pattern (LTP), LEBP and local binary pattern (LBP). © 2012 ACM.